US2626938A - Acidic fractional ester of a polycarboxylic acid with oxypropylated allyl starch - Google Patents

Description

Patented Jan. 27, 1953 ACIDIC FRACTIONAL ESTER OF A POLYCAR- BOXYLIC ACID WITH OXYPROPYLATED ALLYL STARCH Melvin De Groote, University City, Mo., assignor to Petrolite Corporation, a corporation of Delaware No Drawing. Application May 14, 1951, Serial No. 226,325

8 Claims.

.a process for breaking petroleum emulsions of the water-in-oil type characterized by subjecting the emulsion to the action of a demulsifier including hydrophile synthetic products; said hydrophile synthetic products being acidic fractional esters obtained by reaction between (A) polycarboxy acids, and (B) high molal oxypropylation derivatives obtained by the oxypropylation of a member selected from the class consisting of organic solvent-soluble allyl starch and polymerized allyl starch; said oxypropyla tion involving to 50 parts by weight of propylene oxide per unit weight of the allyl starch derivative; and with the final proviso that the ratio of (A) to (B) be one mole of (A) for each hydroxyl radical present in (B).

Although the herein described products have a number of industrial applications, they are of particular value for resolving petroleum emulsions of the water-in-oil type that are commonly referred to as cut oil, roily oil, emulsified oil, etc., and which comprise fine droplets of naturally-occurring waters or brines dispersed in a more or less permanent state throughout the oil which constitutes the continuous phase of the emulsion. This specific application is described and claimed in my co-pending application, Serial No. 226,324, filed May 14, 1951.

The new products are useful as wetting, detergent and leveling agents in the laundry, textile and dyeing industries; as wetting agents and detergents in the acid washing of building stone and brick; as wetting agents and spreaders in the application of asphalt in road building and the like; as a flotation reagent in the flotation separation of various aqueous suspensions containing negatively charged particles, such as sewage, coal washing waste water, and various trade wastes and the like; as germicides, insecticides, emulsifying agents, as for example, for cosmetics, spray oils, water-repellent textile finishes; as lubricants, etc.

For purpose of convenience, what is said hereinafter will be divided into four parts:

Part 1 will be concerned with a brief description of allyl starch and polymerized allyl starch;

Part 2 will be concerned with the oxypropylation of allyl starch and polymerized allyl starch;

Part 3 will be concerned with the preparation of the esters from the oxypropylated derivatives; and

Part 4 will be concerned with derivatives valuable for various purposes including demulsification but not specifically claimed in the instant application.

PART 1 Allyl starch is the name commonly applied to the allyl ether of starch. Allyl ethers of starch are well known and have been described in the literature. See the article entitled Allyl Ether of Starch. Preparation and Industrial Possibilities, by Nichols, Jr., Hamilton, Smith, and Yanovsky. (Industrial and Engineering Chemistry, volume 37, No. 2, February 1945, page 201.)

At least one company, General Mills, Inc., Minneapolis, Minnesota, produces allyl starch commercially. For a complete description of allyl starch see General Mills, Inc., New Product Data Sheet, Revision I, December 15, 1949.

The number of allyl groups introduced per glucose unit vary somewhat but on the average, probably 2.5 allyl groups per glucose unit is high. In the manufacture of allyl starch the glucoses can be purified by dissolving in acetone, filtering oil the small amount of unchanged or lowly substituted starch and precipitating with water. Allyl starches, as one would suspect, are readily soluble in semi-polar or oxygenated solvents such as the alcohols, dipropyleneglycol, ketones, ether alcohols, ester alcohols, etc., and particularly in the nonoxygenated or nonhydroxylated compounds if a small amount of isobutanol is added. They are also generally soluble in all halogenated compounds except carbontetrachloride; for instance, one can usually dissolve 2 grams of allyl starch in 20 grams of suitable solvent without any difiiculty. This is true in regard to benzene. For convenience, the allyl starches employed are referred to as organic solvent-soluble allyl starches or as water-insoluble allyl starches.

The manufacture of commercially available allyl starch states as follows:

Allyl starch is soluble in alcohols, ketones, esters, halogenated hydrocarbons, nitroparafiins, ethers, glycols (in some instances the addition of small amounts of butyl cellosolve or butanol is required to provide good solubility), and in aromatic hydrocarbons provided some hydrogenbond forming solvent such as isobutanol is present. It is insoluble in aliphatic hydrocarbons and turpentine.

Commercial use of allyl starch is largely in the field of coatings, or similar materials, such as bronzing liquids, thermosetting adhesives, overprint and finishing varnish, printing ink vehicles, and the like. In numerous instances the industrial application depends upon the case with allyl starch polymerizes. This fact has been noted in the literature as, for example, in the articles pre- "of diphenyl ether and xylene.

viously cited. This allyl ether of starch insolubilizes with greater ease than the comparatively easier oxidizable allyl ethers of simpler carbohydrates.

Indeed, the manufacturer of commercial allyl starch states as follows:

Solution stability.nce allyl starch is dissolved in a solvent it is protected from air and the solutions may be stored for indefinite periods without danger of gelation. Solutions containing driers may also be stored, provided the container is kept filled in order to displace aira In other words the solution of allyl'starch oxidizes readily by mere exposure to air in the presence or absenceof a catalyst. In theabsence of a-catalyst polymerization takes place -.by simply blowing in the manner, for example, that castor oil is blown at 100 C. or somewhat higher. At a lower temperature .polymerizationtakes place if oneblows in presence of 0.05% of cobalt ..(based on the weight of solids) and in the formof cobalt octoate or cobalt naphthenate. This is described in the previously mentioned New Product Data Sheetof General Mills, Inc. In any event one may prepare any suitable solution of the waterinsoluble allyl starch, subject the same to blowing-with air at a comparatively low temperature inpresence of a catalyst as mentioned, and "stop the blowing at a stage short of gelation and thus have a solution of polymerized allyl starch rather than allyl starch. As far as oxypropylation is concerned, as hereinafter described it is immaterial whether one uses allyl starch or polymerized allyl starch. In either eventthe allyl starches must be water-insoluble and organic solvent-soluble. It goes without'saying it is more convenient if the organic solvent is one which does not interfere with subsequent oxypropylation. All of this as far as suitable solutions are concerned,

will be illustrated by subsequent examples.

Preparation of allyl starch solution stored and shipped under water. Solids can also be isolated from the 40% solution which is'normally sold. The water was drained from the solid material and the powder immersed in a mixture The mass was heated under refluxcondenser with a phase-sep- ,arating trap and water eliminated in the usual manner, along-with xylene. When the water was completely eliminated and part of the xylene removed the final product consisted of 8.5 parts of allyl starch and :13 parts :of solvent. The 13 parts of solvent represented 27 of diphenyl ether and 73% xylene. Any other suitable solvent could be used just as satisfactorily.

PART 2 becomes obvious that it is adapted for oxyethylation as well as oxypropylation.

Oxypropylations are conducted under a wide variety of conditions, not only in regard to presence or absence of catalyst, and the kind of cata- 'lyst, but also in regard to the time of reaction,

temperature of reaction, speed of reaction, pressure during reaction, etc. For instance, oxyalkylations can be conducted at temperatures up to .approximately 200 C. with pressures in about the-same range up to about 200 pounds per square inch. They can be conducted also at temperatures approximating the boiling point of water or slightly above, as for example to C. Under such circumstances the pressure will be less than 30 pounds per square inch unless some special procedure is employed as is sometimes the case, to wit, keeping an atmosphere of inert gas such as nitrogen in the vessel during the reaction. Such low temperature-low reaction rate oxypropylations have been described very completely in U. S. Patent No. 2,448,664, to HR. Fife, et al., dated September 7, 1948. Low temperature, low pressure oxypropylations are particularly desirable where the compound being subjected to oxypropylation contains one, twoor three points of reactiononly, such as monohydric alcohols, I glycols and .tr-iols.

Although the word glycol or diol is usually applied to compounds containing .carbon, hydrogen, and oxygen only, yet the nitrogen- .containing compounds .herein .are diols in..the sensethat they are dihydroxylated. Thus, .the conditions which apply to the oxypropylation of certain glycols also apply in this instance.

Since low pressure-low temperature low-reaction speed oxypropylations require considerable time, for .instance, ,1 to .7 .days of 24 hours each propylene oxide (or for that matter ethylene oxide if it is being used) if the pressure gets .beyond a predetermined range, such as 25 to 35 pounds. Otherwise, the equipment is substantially the same as is commonly employedforthis purpose where thepressure ofreactionis higher, speed of reaction is higher, and time of reaction is much shorter. In such instances such automatic controls are not necessarily used.

Thus, in preparing the various examples ;I have found it particularly advantageous to use laboratory equipment or pilotplant which, is designed to permit continuous oxyalkylation whether it be oxypropylation or oxyethylation. With certain obvious changes the equipment .canbe used also to permit oxyalkylation involving the use of .glycide where no pressure is involved except the vapor .pressureof a solvent, if. any, which may have been used as a diluent.

As previously pointed out the method of using propylene oxide is the same as ethylene oxide. This point is emphasized only for the reason that the apparatus is so designed and constructed as to useeither oxide.

approximately gallons and a working pressure of one thousand pounds gauge pressure. This pressure obviously is far beyond any requirement as far as propylene oxide goes unless there is a reaction of explosive violence involved due to accident. The autoclave was equipped with the conventional devices and openings, such as the variable-speed stirrer operating at speeds from R. P. M. to 500 R. P. M.; thermometer well and thermocouple for mechanical thermometer; emptying outlet; pressure gauge, manual vent line; charge hole for initial reactants; at least one connection for introducing the alkylene oxide, such as propylene oxide or ethylene oxide, to the bottom of the autoclave; along with suitable devices for both cooling and heating the autoclave, -such. as a cooling jacket, and, preferably, coils in addition thereto, with the jacket so arranged that it is suitable for heating with steam or cooling with water and further equipped with electrical heating devices. Such autoclaves are, of course, in essence small-scale replicas of the usual conventional autoclave used in oxyalkylation procedures. In some instances in exploratory preparations an autoclave having a smaller capacity, for instance, approximately 3 /2 liters in one case and about 1% gallons in another case, was used.

, Continuous operation, or substantially continuous operation, was achieved by the use of a separate container to hold the alkylene oxide being employed, particularly propylene oxide. In conjunction with the smaller autoclaves, the container consists essentially of a laboratory bomb having a capacity of about one-half gallon, or

somewhat in excess thereof. In some instances a larger bomb was used, to wit, one having a capacity of about one gallon. This bomb was equipped, also, with an inlet for charging and an eductor tube going to the bottom of the container so as to permit discharging of alkylene oxide in the liquid phase to the autoclave. A bomb having a capacity of about pounds was used in connection with the 15-gallon autoclave. Other conventional equipment consists, of course,

of the rupture disc, pressure gauge, sight feed I other usual conventional procedure or addition I which provided greater safety was used, of course, such as safety glass protective screens, etc.

Attentionis directed again to what has been said previously in regard to automatic controls which shut off the propylene oxide in event temperature of reaction passes out of the predetermined range or if pressure in the autoclave passes out of predetermined range.

With this particular arrangement practical ly all oxypropylations become uniform. in that the reaction temperature was held within a few degrees of any selected point, for instance, if C. was selected as the operating temperature the maximum point would be at the most C. or 112 C., and the lower point would be 95 or tion was comparatively slow under such conditions as compared with oxyalkylations at 200 C.

Numerous reactions were conducted in which the time varied from one day to two days for completion of the final series. In some instances the reaction took place in considerably less time; for instance, at a single stage the reaction may have been complete in 5, 6 or 7 hours. In the series employed for purpose of illustration subsequently, the minimum period of time was 6 hours and the maximum 8 hours. Actually, where an oxypropylation is indicated as being complete in 6 hours it may have been complete in a considerably shorter period of time in light of the automatic equipment employed. This applies,

. also, to other periods of reaction, for instance,

possibly 98 C. Similarly, the pressure was held at approximately 30 pounds within a 5-pound variation one way or the other, but might drop to practically zero, especially where no solvent such as xylene is employed. The speed of reaclonger or shorter. The automatic devices continue stirring for the predetermined period of time even though reaction may have been complete earlier. In the addition of propylene oxide, in the autoclave equipment as far as possible the valves were set so all the propylene oxide if fed continuously would be added at a rate so that the predetermined amount would react within the first 15 hours of the 24-hour period or two-thirds of any shorter period. This meant that if the reaction was interrupted automatically for a period of time for pressure to drop or temperature to drop the predetermined amount of oxide would still be added in most instances well within the predetermined time period. Sometimes where the addition was a comparatively small amount in a 10-hour period there would be unquestionable speeding up of the reaction, by simply repeating the examples and using 3, 4, or 5 hours instead of 10 hours.

When operating at a comparatively high temperature, for instance, between 150 to 200 0., an unreacted alkylene oxide such as propylene oxide, makes its presence felt in the increase in pressure 'or the consistency of a higher pressure. However, at a low enough temperature it may happen that the propylene oxide goes in as a liquid. If so, and if it remains unreacted there is, of course, an inherent danger and appropriate steps must be taken to safeguard against this possibility; if need be a sample must be withdrawn and examined for unreacted propylene oxide. One obvious procedure, of course, is to oxypropylate at a modestly higher temperature, for instance, at to C. Unreacted oxide affects determination of the acetyl or hydroxyl value of the hydroxylated compound obtained.

The higher the molecular weight of the compound, i. e., towards the latter stages of reaction, the longer the time required to add a given amount of oxide. One possible explanation is that the molecule, being larger, the opportunity for random reaction is decreased. Inversely, the lower the molecular weight the faster the reaction takes place. For this reason, sometimes at least, increasing the concentration of the catalyst does not appreciably speed up the reaction, particularly when the product subjected to oxyalkylation has a comparatively high molecular weight.

However, as has been pointed out previously, operating at a low pressure and a low temperature even in large scale operations as much as a week or ten days time may lapse to obtain some of the higher molecular weight derivatives from monohydric or dihydric materials.

In a number of operations the counterbalance scale or dial scale holding the propylene oxide bomb was so set that when the predetermined amount of propylene oxide had passed into the reaction the-scale movement 'through .a time :op-

eratingdevice was setior either onepto two :hours -so-that reaction continued for :1 to 3 hoursafter the final addition-of the last propylene oxide and thereafter the operation was shut down. This '8 three -,s,ubseguent .examples, to wit, in Examples 2a 3a and fla. It may be well to point out that .thecomparatively low pressure was due :to the fact, in part '5 at least, that there was a sizeable concentration particular device is particularly suitable -.for use of catalyst in all four stages of oxypropylation. onlarger equipment than laboratory sizeauto- I The rate of addition of propylene oxide as olaves, to wit, on semi-pilot plant or pilot plant above indicated was comparatively slow. The .size, as well as on large scale size. This final initial introduction of propylene oxide was not stirring period is intended to avoid the presence '10 started .until the heating devices .had raised .the of unreacted oxide. temperature well above the boiling point of In this sort of operation, of course, the temwater, for instance, toabout 240 F. At the .perature range was controlled automatically by completion of the-reaction a sample as taken either use of cooling water, :steam, or .electrical and oxypropylation proceeded was in Example heat, so asto raise or lower the temperature. 1-5 2a,immediatelyiollowing.

The pressuring of the propylene oxide into the Example 2a reaction vessel was also automatllc insofar that SW75 pounds of the reaction mass identified the Stream was for a 5 w muous as Example'la, preceding, and equivalent to .74 lmlm'whlch shut 9 1 case i g g z fi I pound of allyl starch, 38.2 pounds of propylene P as pretvlous. y t A oxide, .85 pound of causticsoda, and 11.3-pounds 8 8 eslgn cons women 8 w of solvent were subjected to furtheroxypropylaconventional lpcludmg the gases check valves tion without the addition of any more catalyst. I enme eqmpment far as I am i t 5 in the same manner as described in Example 10;, least two .firms, and possibly three, speclallzeln preceding autoclave eqPipment such I have m The amount of oxide "added was 42.75 pounds. "the q and furnisheqmp' The time of addition was 7 hours. The rate of or tllus m .snnguatly .pllot "plant addition was about 6 or '7 pounds per hour. At 'equlpment lslavzflflable' T P 1s slmply made the end of the reactionperiod part of the reacprfacautwn the duiectlolt'of Safety Oxytion mass was withdrawn and the remainder p partlculafly involvmg ethylene subjected to further oxypropylation-as described glyclde propylfne shquld mt be in Example 311, immediately following.

conducted except in equlpment specifically :de- "signed for the purpose. Ezamp le Example 1a 62.75 pounds of reaction mass identified as Example 2a, preceding, and equivalent to 4.62

Thestarting material was a mixture of allyl pounds of allyl starch, 50.55 pounds of propylene starch-and solvent as previously described. More oxide, .53 pound of caustic soda, and 7.05p0unds specifically, the mixture consisted of 8.5 pounds of solvent, were subjected to further oxypropylaof :allylstarchof a commercial grade, 3..5 .pounds tion Wi h ut he a di on f any m re a aly ofdiphenyl ether, and 9.5 pounds of xylene. Th in the same manner as employed in Examples 1a particular autoclave employed was one with a and 2a, p e ng. The amount of oxide added capacity of about 15 :gallons, or on the average was 44.25 pounds. The time requ ed to add of about 125 pounds ofthe reaction mass. The the oxide was .8 hours. The oxide was added at initial charge was as previously indicated, and the ratesof about 6 pounds p 1101111 t e en including one pound of callstic'soda. The reacf h r n p r d p f the reaction mass tionpot .was flushed out with nitrogen, the autowas Withdrawn and the remainder Subjected to clave sealed, and theautomatic devices adjusted further QXYPI'OP.ylatiolr1 in the 11134111181 described 'for injecting 43.75pounds of propylene :oxide in m Example jouowmigapproximately a '6--hour period. The oxide was Eimmple 4a :added at the rate of about 9 or 10 pounds per 5U -61.75'pounds of the reaction mass identified-as hour. The pressure regulator was set for a Example P c d n We e S jec d o further maximum of 15 t 20 pounds vSquare inch, oxypropylation without the addition of anymore In othersimflar experiments Iha ve used-a ome Patailyst, Same procedure as noted What'higher pressure, for instance, a maximum m Examples through p spressure .of 35 to 37 pounds per square inch. 9 of oxlde r d w pounds; h -However, in this particular instance the .presi g m r The r. sure did .not actually reach .a maximum of .over i g 2 5 Per he procedme as as oxyplopylatlon was concerned was the 20 pounds per square inch. .1 have found no dlf- .same as in preceding examples .fi cultyin conducting this reaction under condi- What has been Said heren'l is presented in houses descrlbed at thls comparatively low prestabular for in Table 1 immediately following, re- The temp r r mpl y w 6 with some added information as to molecular tio as for a temperature and weight and as to solubility of the reaction prod- .pressure were concerned are identical in the uct in water,xylene, and kerosene.

' TABLE 1 Composition Before Composition at End i Max I iii H. .o Oxide. Cata- 5; Theo. .11. o .Oxide Oata- 1 11? amt, 1411112., lyst, Amt M01. Amt, Amt, lyst, gi

lbs. lbs. lbs. wt. lbs lbs. lbs.

1o 8.50 1.0 13.0 1,240 8.50 43.75 1.0 13.0 1.008 250 200 15-20 -0' 7.40 30.2 .85 11.3 2,545 7.40 80.95 .85 11.3 1, 500 250-200 15-2 7 4.02 50.55 .53 7. 05 4,790 4.02 94.80 .53 7. 05 2, 34s 250 200 15 20 s 2.00 54.72 .31 -4. 00 0,540 2.00 7577 .51 4:00 2,805 250-200 15 20 5% ated compounds.

Example 1a was emulsifiable in water, insoluble in xylene, and insoluble in kerosene; Example 2a was emulsifiable in water, soluble in xylene but insoluble in kerosene; Example 3a. was emulsifiable to insoluble in water, soluble in xylene, and dispersible in kerosene; and Example 4a was emulsifiable to insoluble in water, soluble in xylene and also in kerosene.

In the above table the molecular weight figures are, of course, open to speculation. Since the molecular Weight of allyl starch itself is unknown the nearest approach to a molecular weight relationship depends on the glucose unit as a basis of comparison. Needless to say, the allyl ethe-rs represent a somewhat greater molecular weight than the corresponding glucose unit. For purpose of the preceding table I have used a figure for the unit of 2'22. As is pointed out elsewhere, oxypropylation is a rather complicated procedure, particularly when polyhydric materials are employed, especially in such instances where more than 2 hydroxyls are available per unit or per molecule. In any event, it is tobe noted that the initial allyl starch was combined with to 50 times its weight of propylene oxide, based on the assumption of completeness of reaction.

In other series I have added more catalyst and continued to oxypropylate until the molecular weight range was approximately twice that, i. e., a theoretical molecular weight range of 10,000 to 12,000 where the allyl starch represented approximately 1% or slightly more of the final reaction mass. In such instances, however, the hydroxyl molecular weight rose somewhat less, to a maximum of 4,000 to 4,500.

These products were invariably kerosene-soluble as well as being xylene-soluble and insoluble in water.

The final product at the end of the oxypropylation step was amber or pale amber, or of a pale straw color in some instances. This was more or less characteristic of all the various oxypropylations products in the various stages. These products were, of course, slightly alkaline due to the residual caustic soda and also due to the basic nitrogen atom. The residual basicity due to the catalyst, of course, would be the same if sodium methylate had been used.

Speaking of insolubility in water or solubility in kerosene such solubility test can be made simply by shaking small amounts of the materials in a test tube with water, for instance, using 1% to 5% approximately based on the amount of water present.

I ,Needless to say, there is no complete conversion of propylene oxide into the desired hydroxyl- This is indicated by the fact that the theoretical molecular weight based on a statistical average is greater than the molecular weight calculated byusual methods on basis of acetyl or hydroxyl value. Actually, there is no completely satisfactory method for determining molecular weights of these types of compounds with a high degree of accuracy when the molecular weight exceeds 2,000. In some instances the acetyl value of hydroxyl value serves as satisfactorily as an index to the molecular weight as any other procedure, subject to the above limitations, and especially in the higher molecular weight range.

PART 3 As previously pointed out the present inventions concerned with acidic esters obtained from the oxypropylated derivatives described in Part particularly tricarboxy acids like citric and dicarboxy acids such as adipic acid, phthalic acid, or an'hydride, succinic acid, di-gly-colic acid, sebacic acid, azelaic acid, aconitic acid, maleic acid or anhydride, citraconic acid or anhydrde, maleic acid or anhydride adducts as obtained by the Diels-Alder reaction from products such as maleic anhydride, and cycl-opentadiene. Such acids should be heat stable so they are not decomposed during esterification. They may contain as many as 36 carbon atoms as, for example, the acids obtained by dimerization of unsaturated fatty acids, unsaturated monocarboxy fatty acids, or unsaturated monocarboxy acids having 18 carbon atoms. Reference to the acid in the hereto appended claims obviously includes the anhydrides or any other obvious equivalents. My preference, however, is to use polycarboxy acids having not over 8 carbon atoms.

The production of esters including acid esters (fractional esters) from polycarboxy acids and glycols or other hydroxylated compounds is well known. Needless to say, various compounds may be used such as the low molal ester, the anhydride, the acyl chloride, etc. However, for purpose of economy it is customary to use either the acid or the anhydride. A conventional procedure is employed. On a laboratory scale one can employ a resin pot of the kind described in U. S. Patent No. 2,499,370, dated March '7, 1950, to De Groote & Keiser, and particularly with one more opening to permit the use of a porous spreader if hydrochloric acid gas is to be used as a catalyst. Such device or absorption spreader consists of minute Alundum thimbles which are connected to a glass tube. One can add a sulfonic acid such as paratoluene sulfonic acid as a catalyst. There is some objection to this because in some instances there is some evidence that this acid catalyst tends to decompose or rearrange heat oxypropylated compounds, and particularly likely to: do so if the esterification temperature is too high. In the case of polycarboxy acids such as diglycollic acid, which is strongly acidic there is no need to add any catalyst. The use of hydrochloric acid gas has one advantage over paratoluene sulfonic acid and that is that at the end of the reaction it can be removed by flushing out with nitrogen, whereas there is no reasonably convenient means available of removing the paratoluene sulfonic acid or other sulfonic acid employed. If hydrochloric acid is employed one need only pass the gas through at an exceedingly slow rate so as to keep the reaction mass acidic. Only a trace of acid need be present. I have employed hydrochloric acid gas or the aqueous acid itself to eliminate the initial basic material. My preference, however, is to use no catalyst whatsoever.v

The products obtained in Part 2 preceding may contain a basic catalyst. As a general procedure I have added an amount of half-concentrated hydrochloric acid considerably in excess of what is required to neutralize the residual catalyst. The mixture is shaken thoroughly and allowed to stand overnight. It is then filtered and refluxed with the xylene present until the water can be separated in a phase-separating trap. As soon as the product is substantially free from water the distillation stops. This preliminary step can be carried out in the flask to be used for esterification. If there is any further deposition of sodium chloride during the reflux stage needless to say a second filtration may be required. In any event the neutral or slightly acidic solution of the oxypropylated derivatives described in Part 2 is then diluted further with sufiicient xylene, decalin, petroleum solvent, or the like, so that one has obtained approximately a 45% solution. To this solution there is added a polycarboxylated reactant as previously described, such as phthalic anhydride, succinic acid or anhydride, diglycolic acid, etc. The mixture is refluxed until esterification is complete as indicated by elimination of water or drop in carboxyl value. Needless to say, if one produces a halfester from an anhydride such as phthalic anhydride, no water is eliminated. However, if it is obtained from diglycolic acid, for example, water is eliminated. All such procedures are conventional and have been so thoroughly described in the literature that further consideration will be limited to a few examples and a comprehensive table.

Other procedures for eliminating the basic residual catalyst, if any, can be employed. For example, the oxyalkylation can be conducted in absence of a solvent or the solvent removed after oxypropylation. Such oxypropylation end-product can then be acidified with just enough concentrated hydrochloric acid to just neutralize the residual basic catalyst. To this product one can then add a small amount of anhydrous scdium sulfate (sufiicient in quantity to take up any water that is present and then subject the mass to centrifugal force so as to eliminate the hydrated sodium sulfate and probably the sodium chloride formed. The clear, somewhat viscous amber colored or pale straw-colored or light amber liquid so obtained may contain a small amount of sodium sulfate or sodium chloride, but, in any event, is perfectly acceptable for esterification in the manner described.

It is to be pointed out that the products here described are not polyesters in, the sense that there is a plurality of both allyl starch radicals and acid radicals; the product is characterized by having only one allyl starch radical.

In some instances and, in fact, in many instances I. have found that in spite of. the dehydration methods employed above that a mere trace of water still comes through and that this mere trace of water certainly interferes with the .acetyl or hydroxyl value determination, at least when the number of conventional procedures are used, and may retard esterification, particularly where there is no sulfonic acid or hydrochloric acid present as a catalyst. Therefore, I have preferred to use the following procedure: I have employed about 200 grams of the polyhydroxyl'ated compound as described in Part 2, preceding; I have added about 60- grains of benzene, and then refluxed this mixture in the glass resin pot using a phase-separating trap until the benzene carried out all the water present as water of solution or the equivalent. Ordinarily this refluxing temperature is apt to be in the neighborhood of 130 to possibly 150 C. When all this water or moisture has been removed I also withdraw approximately grams or a little less benzene and then add the required amount of the carboxy reactant and also about 150 grams of a high boiling aromatic petroleum solvent. These solvents are sold by various oil refineries and, as far as solvent effect, act as if they were almost completely aromatic in character. Typical distillation data in the particular 12 type I have employed and found very satisfactory is the following:

I. B. P., 142 C. ml., 242 C. 5 ml., 200 C. m1., 244 C'. 10 ml., 209 C. ml., 248 C. 15 ml., 215 C. ml., 252 C. 20 ml., 216 C. 1111., 252 C. 25 ml., 220 C. '75 1111., 260 C. 30 m1., 225 C. ml., 264 C. 35 1111., 230 C. 1111., 270 C.

40 ml., 234 C. m1., 280 C. 45 ml., 237 C. 1111., 307 C.

After this material is added, refluxing is continued, and, of course, is at a high temperature, to wit, about to C. If the carboxy reactant is an anhydride needless to say no water of reaction appears; if the carboxy reactant is an acid, water of reaction should appear and should be eliminated at the above reaction temperature. If it is not eliminated I simply separate out another 10 to 20 cc. of benzene by means of the phase-separating trap and thus raise the temperature to or 190 C., or even to 200 C., if need be. My preference is not to go above 200 C. This particular example is illustrative only. Other proportions may be employed.

The use of such solvent is extremely satisfactory provided one does not attempt to remove the solvent subsequently except by vacuum distillation and provided there is no objection to a little residue. Actually, when these materials are used for a purpose such as demulsification the solvent might just as well be allowed to remain. If the solvent is to be removed by distillation, and particularly by vacuum distillation, then the high-boiling aromatic petroleum solvent might well be replaced by some more expensive solvent, such as decalin or an alkylated decalin which has a rather definite or close range boiling point. The removal of the solvent, of course,

benzene mixture.

geneous solution.

ether and methanol. anol is not present during the esterification procis purely conventional procedure and requires no elaboration.

If the products or compounds subjected to oxypropylation tend to stay water-soluble to any significant degree as, for example, when polyamines are oxypropylated, one may find that after the water is eliminated the freshly formed ester is not soluble in the hydrocarbon solvent such as xylene, or the petroleum solvent- An addition of a semi-polar solvent, such as methanol or the diethylether of diethyleneglycol can be conveniently employed. Of course, the immiscible solvent, such as xylene, can be removed and the product dissolved in methanol, ethanol, propanol, or the like, as a matter of convenience.

Actually the oxypropylated derivatives of allyl starch or polymerized allyl starch when subjected to esterification with polyhydric acids show a tendency towards gelation or cross-linking, or at least show a tendency towards a conan oxygenated solvent in addition to a hydrocarbon solvent in order to obtain a clear homo- This has been referred to previously but reference is made specifically to Table 2 wherein the solvent employed is a mixture of benzene, xylene, diethylene glycol diethyl- Needless to say, the methess but isadded at the end.

' Inthese experiments the total amount ofs olth yl ther Of diethylene glycol.

vent is indicated. The solvent used represents approximately 65% of diethylether of diethylene glycol and 15% of benzene. Over and above this there-was also present the xylene and diphenyl ether which was used as a solvent in the oxypropylation step as described in Part 2, preceding; At the completion of the esterification step in order to obtain a homogeneous solution a'small amount of methanol was added. The amount of methanol was usually 5%, 6%, or 7% of the'nonhydroxylated solvents previously present.

In subsequent Table 2 reference to amount of solvent is the total solvent, and reference to diethyl carbitol is, of course, reference to di- These percen tages are not critical and may be varied for convenience. Other solvents could, of course, be substituted as previously pointed out. However, this particular mixture, at least in the subsequent examples, has worked satisfactorily. The esterification step, as has been pointed out, is conventional and the data are summarized in Tables 2 and 3, following.

The procedure for manufacturing the esters has been illustrated by preceding examples. If for any reason reaction does not take place in a manner that is acceptable, attention should be directed to the following details: (a) Recheck the hydroxyl or acetyl value of the oxypropylated primary amines of the kind specified and use a stoichiometrically equivalent amount of acid; (b) if the reaction does not proceed with reasonable speed either raise the temperature indicated or else extend the period of time up to 12 or 16 hours if need 'be; (c) if necessary, use /2% of paratoluene sulfonic acid or some other acid as a catalyst; (d) if the esterification does not produce a clear product a checkshould be made V TABLE 2 'Theo M01. Amt. of F No Ex. Theo. Actual Wt. Amt. of Poly- A No. of M. W (am-W1 Hy- Based Hyd. Polycarboxy carboxy Oxy. of V drox. on Cmpd Reactant Re- 5 er Cmpd H. C 0 Value Actual (grs.) aetant H. V (grs.)

la 1, 240 154 1,098 109. 8 Diglycolie Acid 40. 2 1a 1, 240 135 154 1, 098 109. 8 Oxalic Acid 37, 8 1a 1, 240 135 154 1, 098 109. 8 Aconitic Acid 52. 2 1a 1, 240 135 154 1,098 109. 8 Adlpic Acid 43. 8 1a 1, 240 135 154 l, 098 109. 8 Phthalic Anhyd 44. 4 1a 1, 240 135 154 1, 098 109. 8 Maleic Anhyd l 29. 4 2a 2, 545 66. 2 112 1, 506 150. 6 Diglycolic Acid 40. 2 2a 2, 545 66. 2 112 1, 506 150. 6 Oxalic Acid 37. 8 2a 2, 545 66. 2 112 1, 506 150. 6 Aconitic Acid 52. 2 2a 2, 545 66. 2 112 1, 506 150. 6 Adiflic Acid 43. 8 2a 2, 545 66. 2 112 1, 506 150. 6 Phthalic Anln 44. 4 2a 2, 545 66. 2 112 1, 506 150. 6 Maleic Anhyd 29. 4 3a 4,790 35. 2 71. 8 2, 343 117 Diglycolic A01 20. 1 3a 4, 790 35. 2 71. 8 2, 343 117 OX 10 Ac l8. 9 3a 4, 790 35. 2 71. 8 2, 343 117 Aconitic A01 26. 1 3a 4, 790 35. 2 71.8 2, 343 117 Adipic Acid 21.9 3:! 4, 790 35. 2 71. 8 2, 343 117 Phthalio Anhyd 22. 2 3a 4, 790 35. 2 71. 8 2, 343 117 Malcic Anhyd l4. 7 4a 6, 540 25. 8 60. 0 2, 805 Diglycolic Acid 20. 1 4a 6, 540 25. 8 60.0 2, 805 140 18. 9 4a 6, 540 25. 8 60. 0 2, 805 140 26. 1 4a 6, 540 25. S 60. 0 2, 805 140 21. 0 4a 6, 540 25. 8 60. 0 2, 805 140 22. 2 4a 6, 540 25. 8 60. 0 2, 805 140 Maleic Anhyd' 14. 7

TABLE 3 Ester- Time of Ex. N0. gfj ifica- Ester- Water of Acid Solvent vent Ttion itfica- (0111): Ester emp. 1011 C0. (m) 0. (hrs.)

16. Q L Benzene, xylene, Diethyl 143. 4 158 7 6. 6 Carbitol methanol.

Even und r the mo tcarefully controlled conditi s f 'oxypr ylation inv lvin comp v l low temperatures and long time of reaction there are formed certain compounds whose compositions are still obscure. Such side reaction products can contribute a substantial proportion of the final cogeneric reaction mixture. Various su estions have been made as to the nature of these compounds, such as being cyclic polymers .Of propylene oxide, dehydration products with the ap-[ pearance 10f a vinyl radical, or isomers of propylene oxide or derivatives thereof, 1. e., of an aldehyde, ketone, or allyl alcohol. In some instances an attempt to react the stoichiometric amount of a polycarboxy acid with the oxypropylated derivative results inan excess of the carboxylated reactant for the reason that apparently under conditions of reaction less reactive hydroxyl radicals are presentthan indicated by the hydroxyl value. Under such circumstances'there is simply aresidue of the carboxylic reactant which can be removed by filtration or, if desired, the esterification procedure can be repeated using an appropriately reduced ratio of carboxylic reactant.

Even the determination of the hydroxyl value and conventional procedure leaves much to be desired due either to the cogeneric materials previously referred to, or for that matter, the presence of any inorganic salts or propylene oxide. Obviously this oxide should be eliminated.

The solvent employed, if any, can be removed from the finished ester by distillation and particularly vacuum distillation. The final products or liquids are generally pale reddish amber to reddish amber in color, and show moderate viscosity. They can be bleached with bleaching clays, filtering chars, and the like. However, for the purpose of demulsification or the like color is not a factor and decolorlzation is not justified.

In the above instances I have permitted the solvents to remain present in the final reaction mass. In other instances I have followed the same procedure using decalin or a mixture of decalin or benzene in the same manner and. ultimately removed all the solvents by vacuum distillation. Appearance of the final productsis much the same as the diols before esterification and in some instances was straw, pale amber, or dark amber in color and had a more reddish cast and perhaps was somewhat more viscous.

PART 4 As pointed out previously, the final product obtained is a fractional ester having free carboxyl radicals. Such product can be used as an intermediate for conversion into other derivatives which are effective for various purposes, such as the braking of petroleum emulsions of the kind herein described. For instance, such product can be neutralized with an amide so as to increase its Water-solubility such as triethanolamine, triproparticularly true where surface-active,gnaterials are of value and especially water-in-oil emulsions.

Having thus described my invention, what ,I claim as new and desire to secure by Letters Pat..- ent, is:

1. A hydrophile synthetic product which is an acidic fractional ester of (A) a polycarbQXy acid, and (B) a polyhydroxycompound selected from the class consisting of oxypropylated, organic solvent-soluble, allyl starch and .oxypropylated polymerized allyl starch; said polyhydroxycompound containing in combination, 5 to parts by weight of propylene oxide per unit weight of the allyl starch derivative, with the proviso that the ratio of (A) to (B) be one mole of (A) for each hydroxyl radical present in (B) 2. A product as in claim 1 in which the polycarboxy acid is a dicarboxy. acid.

3. A product as in claim 2 in which the dicarboxy acid has less than 8 carbon atoms.

4. The product of claim 2 wherein the dicar-. boxy acid is phthalic acid.

5. The product of claim 2 wherein the di! carboxy acid is maleic acid.

6. The product of claim 2 wherein the dicarboxy acid is succinic acid.

7. The product of claim 2 wherein the dicarboxyacid is citraconic acid.

8. The product of claim 2 whereinthe dicarboxy acid is diglycolic acid.

his MELVIN DE GROOTE. mark in demulsification .pf

Witnesses to mark:

W.,C.ADAMs, I. S. DEGRoorE.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS

Claims (1)

1. A HYDROPHILE SYNTHETIC PRODUCT WHICH IS AN ACIDIC FRACTIONAL ESTER OF (A) A POLYCARBOXY ACID, AND (B) A POLYHYDROXY COMPOUND SELECTED FROM THE CLASS CONSISTING OF OXYPROPYLATED, ORGANIC SOLVENT-SOLUBLE, ALLYL STARCH AND OXYPROPYLATED POLYMERIZED ALLYL STARCH; SAID POLYHYDROXY COMPOUND CONTAINING IN COMBINATION, 5 TO 50 PARTS BY WEIGHT OF PROPYLENE OXIDE PER UNIT WEIGHT OF THE ALLYL STARCH DERIVATIVE, WITH THE PROVISO THAT THE RATIO OF (A) TO (B) BE ONE MOLE OF (A) FOR EACH HYDROXYL RADICAL PRESENT IN (B).